According to the 2025 annual figures (final report) of the Japan Tourism Agency’s “Travel and Tourism Consumption Trend Survey,” released on 30 April 2026, 45.8% of domestic trips taken by Japanese residents are day trips. Spend per person per day trip is about ¥20,000, against ¥72,412 for an overnight trip. For the same visitor, whether they stay the night or head home changes what the region earns by roughly ¥52,400. This article quantifies, prefecture by prefecture, how thick this “demand that visits but does not stay” is on the 2025 final figures, cross-references it with our internal room stock, estimated occupancy and estimated settled ADR, and identifies the 20 prefectures that combine heavy day-trip demand with spare room capacity. It then examines how to build value into the evening, drawing on dinner premiums and day-of-week supply-demand data.

45.8% of All Trips Are Day Trips — a ¥52,400 Gap per Person Between Staying and Going Home

Start with the national picture. According to the Japan Tourism Agency’s “Travel and Tourism Consumption Trend Survey,” 2025 annual figures (final, released 30 April 2026), Japanese domestic travel in 2025 comprised 299.97 million overnight person-trips and 253.16 million day-trip person-trips. Of the 553.13 million total, 45.8% involved no overnight stay.

In monetary terms the gap is starker. Spend per person per trip was ¥72,412 for overnight travel and ¥20,000 for day trips — a difference of ¥52,412, or 3.62x. Limited to leisure and recreation purposes, the ratio widens to 4.03x (¥82,270 against ¥20,416). Japanese domestic travel spend reached ¥26.7845 trillion (+6.5% year on year, +22.1% versus 2019), close to a record high, but it splits into ¥21.7211 trillion from overnight travel against ¥5.0634 trillion from day trips: day trips are 45.8% of people but only 18.9% of the money.

Source: Compiled by the HotelBank Editorial Team from the Japan Tourism Agency, “Travel and Tourism Consumption Trend Survey,” 2025 annual figures (final)

Of the ¥72,412 spent on an overnight trip, only ¥18,205 (25.1%) is accommodation. The rest is transport (¥18,447), food and drink (¥9,721), shopping (¥8,366) and entertainment services (¥4,624). One additional night is room revenue for the hotel and, at the same time, evening dining and next-morning spend for the region. That is why accommodation operators and DMOs have reason to look at the same numbers.

Prefecture-Level Day-Trip Ratios Are Not Published — the Origin-Basis and Destination-Basis Gap

Before breaking this down by prefecture, the granularity of the statistics needs checking. The national figures in the “Travel and Tourism Consumption Trend Survey” come from a sample survey of residents — origin-basis (place of residence) traveller statistics — and carry the overnight/day-trip split. The [Reference] prefectural tabulation the same survey publishes separately is destination-basis, showing visitor counts and spend per visitor for each of the 47 prefectures. That reference table, however, has no overnight/day-trip breakdown. Its own notes state that the “visitor count” in the table includes both overnight and day-trip visitors.

A prefecture-level day-trip ratio therefore cannot be taken directly from this survey. This article instead constructs a proxy metric that combines two public statistics of differing granularity.

This is a proxy metric, and its constraints should be stated. First, the numerator is person-nights and the denominator person-trips, so the units differ. A value of “0.35 nights” is a mechanical quotient meaning an average of 0.35 nights per visit, not an overnight-guest ratio as such. Second, the two surveys differ in sample design and in population. Third, the consumption survey is a sample survey, with prefectural sample sizes ranging from 152 (Fukui) to 1,393 (Hokkaido). Because single-year prefectural values carry sampling error, this article works primarily from the 2025 cross-section and uses a pooled 2024-2025 two-year value as a robustness check. Fourth, the accommodation statistics in the numerator cover all operators regardless of OTA listing, whereas the room stock, ADR and estimated occupancy used later are limited to properties whose operation can be confirmed on OTAs.

The metric is still worth using, because no other combination of published statistics allows all 47 prefectures to be lined up on the same definition in 2025, the latest finalised year. Analysis using the common-standard visitor-arrival statistics can obtain the overnight-guest ratio directly, but the national compilation of those statistics ended with the 2022 annual figures and covers only 20 prefectures. Those visitor-arrival-based overnight ratios by prefecture are set out in Gifu 8.3% vs Wakayama 29.7%: Overnight Stay Rates in 20 Prefectures, and the two analyses complement each other.

A 6.9-Fold Spread — from Saitama 0.33 and Nara 0.35 to Okinawa 2.26

Calculating the overnight conversion ratio for all 47 prefectures gives a national total of 0.746 nights per visit and a 47-prefecture median of 0.711. The lowest is Saitama at 0.328, followed by Nara at 0.353 and Saga at 0.432; the highest is Okinawa at 2.256, a 6.9-fold spread between the extremes. Ishikawa at 1.150 and Hokkaido at 1.083 also exceed 1.

The chart below plots the overnight conversion ratio on the horizontal axis against room occupancy from the “Accommodation Survey” (2025 final figures) on the vertical axis, with bubble size representing room stock. The dashed lines mark the median of 0.711 and the national occupancy rate of 61.6%. Chart labels are in Japanese, as in the original edition.

Source: Compiled by the HotelBank Editorial Team from the Japan Tourism Agency, “Travel and Tourism Consumption Trend Survey,” 2025 annual figures (final) / Japan Tourism Agency, “Accommodation Survey,” 2025 annual final figures / MetroEngines Research

The lower-left quadrant — below the median conversion ratio and below the national average occupancy — is what this article calls the set of prefectures with the most room to convert. Twenty prefectures sit there. At the right edge are Okinawa (2.256, 60.3%) and Hokkaido (1.084, 61.2%), which carry the heaviest overnight demand per visit in the country while their room occupancy sits marginally below the national average. The upper-right quadrant, high on both conversion and occupancy, holds the large metropolitan prefectures: Tokyo (0.841, 76.3%), Osaka (0.737, 78.6%) and Fukuoka (0.833, 72.1%). Saitama, Aichi and Chiba fall upper-left — heavy day-trip demand, but occupancy already above the national average. Because their prescription differs, they are treated separately below.

It is worth checking whether the proxy metric carries real economic meaning. Plotting the overnight conversion ratio against spend per visitor (also from the prefectural tabulation) gives a correlation coefficient of r = 0.841 (N=47). That figure is, however, pulled strongly by the outlier Okinawa (2.256 nights, ¥89,880). Recalculating across the 46 prefectures excluding Okinawa gives r = 0.523, and excluding Hokkaido, Kagoshima and Ishikawa as well leaves r = 0.346 across 43. The direction is consistently positive, but its strength depends on which prefectures are included, and should be read with that discount.

Source: Compiled by the HotelBank Editorial Team from the Japan Tourism Agency, “Travel and Tourism Consumption Trend Survey,” 2025 annual figures (final), [Reference] prefectural tabulation / Japan Tourism Agency, “Accommodation Survey,” 2025 annual final figures (N=47 prefectures)

The 20 Prefectures with the Most Room to Convert — 7,114 Properties and 317,787 Rooms

Twenty prefectures meet both conditions: an overnight conversion ratio below the 47-prefecture median of 0.711, and room occupancy below the national average of 61.6%. On the basis of properties whose operation MetroEngines Research can confirm, these 20 prefectures hold 7,114 properties and 317,787 rooms in total.

Table 2 | In ascending order of overnight conversion ratio. Sample size is the prefectural sample count in the prefectural tabulation of the “Travel and Tourism Consumption Trend Survey.”

Source: Compiled by the HotelBank Editorial Team from the Japan Tourism Agency, “Travel and Tourism Consumption Trend Survey,” 2025 annual figures (final) / Japan Tourism Agency, “Accommodation Survey,” 2025 annual final figures / MetroEngines Research

The lowest conversion ratio is Nara at 0.353. It draws 7.53 million Japanese visitors a year, yet Japanese guest nights stop at 2.66 million person-nights. Room occupancy is 50.9% (2025 final figures, as below), 10.7pt under the national level, and the room stock itself is small at 154 properties and 6,728 rooms. Saga at 0.432 (4.87mn visits, 2.11mn guest nights, 52.4% occupancy), Wakayama at 0.544 (7.50mn, 4.08mn, 48.4%) and Ibaraki at 0.546 (12.75mn, 6.96mn, 54.1%) follow.

Pooling 2024 and 2025 into a two-year value leaves the leading names broadly stable. On the pooled basis the lowest are Nara 0.333, Saitama 0.348, Saga 0.494, Yamaguchi 0.545, Ibaraki 0.550 and Yamanashi 0.557 — and some prefectures move sharply between the two views (Yamaguchi ranks 17th on 2025 alone but 4th on the two-year pool). Single-year values swing more for prefectures with smaller samples, which is best read alongside the sample-size column in Table 2.

Upper Bound of 13.85mn Room-Nights and ¥148.2bn a Year — Absorbable Within 19% of Vacancy

So how much demand would arise if these 20 prefectures reached the median conversion ratio of 0.711 nights per visit? Holding visitor counts constant, we derive the additional guest nights from the gap in conversion ratio, then convert to room-nights by dividing by each prefecture’s guests per occupied room (guest nights ÷ rooms sold, from the “Accommodation Survey”). Nationally, 1.644 guests occupy each room sold. How widely this guests-per-room figure — and therefore the supply-demand headroom — varies across the 47 prefectures is analysed in detail in June 2026 Occupancy 56.2%: Guest-Nights per Room in 47 Prefectures.

Table 3 | Annual vacant room-nights = estimated room stock × 365 × (1 − room occupancy). Estimated room stock is derived as rooms sold ÷ room occupancy ÷ 365. The room-revenue conversion is additional room-nights × estimated settled ADR, and is an indication of an upper bound only.

Source: Compiled by the HotelBank Editorial Team from the Japan Tourism Agency, “Travel and Tourism Consumption Trend Survey,” 2025 annual figures (final) / Japan Tourism Agency, “Accommodation Survey,” 2025 annual final figures / MetroEngines Research

Source: Compiled by the HotelBank Editorial Team from the Japan Tourism Agency, “Travel and Tourism Consumption Trend Survey,” 2025 annual figures (final) / Japan Tourism Agency, “Accommodation Survey,” 2025 annual final figures / MetroEngines Research

Additional room-nights across the 20 prefectures total 13.85 million a year. Against that, the same 20 prefectures carry 72.48 million vacant room-nights annually, so the additional demand fills only 19% of existing vacancy. In most of these prefectures, in other words, the current room stock can absorb it without waiting for new supply. The room-revenue equivalent is roughly ¥148.2 billion, but this is a theoretical ceiling — visitor counts unchanged while the conversion ratio alone rises to the median — not a forecast. Even at one tenth of it, some ¥14.8 billion of additional room revenue would be added a year.

Circumstances differ by prefecture, however. Nara’s absorption rate is 88% and the implied occupancy uplift +43.2pt: taking the full theoretical conversion would push occupancy to around 94%, which does not hold up in practice, so in Nara the expansion of room stock itself becomes part of the discussion alongside demand conversion. Saga (63% absorption, +30.1pt) has a similar character. Yamanashi (12%), Gunma (18%), Shizuoka (2%) and Gifu (10%), by contrast, have ample vacancy and can concentrate on demand-side design.

The upper-bound estimate should be read as a range rather than a single point. Holding visitor counts constant, the table below sets an attainment rate for how much of the gap to the median the 20 prefectures close, and rescales additional room-nights and room revenue proportionally. An attainment rate of 100% corresponds to the theoretical ceiling above.

Table 3-1 | Additional room-nights and room revenue are linear scalings of the 20-prefecture totals in Table 3 (13.85mn room-nights, ¥148.2bn) by the attainment rate. The absorption denominator is the same 20 prefectures’ 72.48 million annual vacant room-nights. Attainment rates are analytical settings and do not indicate probability of realisation.

Source: HotelBank Editorial Team estimates based on Table 3

The room-revenue conversion also depends on the level of estimated settled ADR. Flexing both attainment rate and ADR gives a range of ¥33.3bn to ¥163.0bn a year. A ±10% move in ADR (±¥14.8bn) matters 2.5 times less than a 25-point move in attainment rate (about ¥37.1bn). What drives room revenue is not rate, but how many more people can be persuaded to stay.

Table 3-2 | Flexed on attainment rate and estimated settled ADR against the ¥148.2bn room-revenue conversion in Table 3. The ±10% on estimated settled ADR assumes a uniform shift in the prefectural estimates and does not represent an actual pricing plan.

Source: HotelBank Editorial Team estimates based on Table 3

The Night People Stay Is Concentrated on Saturday — Sunday Estimated Occupancy Runs 10.8pt Lower

So which nights hold that conversion opportunity? Averaging estimated occupancy across the 20 prefectures by day of week (OTA-listed inventory basis, actual dates in June and July 2026) gives 90.3% on Saturday against 79.5% on Sunday, and 80.0% on Monday — Saturday alone stands out. Compared with the 47-prefecture average (Saturday 91.2%, Sunday 81.0%), the 20 prefectures sit 1-2pt lower across the board.

Source: MetroEngines Research; compiled by the HotelBank Editorial Team

Listed prices show the same shape. Average day-of-week listed prices across the 20 prefectures (all-plan average, tax included) run ¥25,705 on Saturday against ¥18,962 on Sunday (74% of Saturday), ¥17,731 on Monday (69%) and ¥19,578 on Friday (76%). Individually, Tottori runs ¥24,745 on Saturday against ¥16,916 on Sunday (1.46x), Saga ¥33,079 against ¥23,889 (1.38x) and Nara ¥25,620 against ¥18,740 (1.37x) — in each case Saturday alone jumps.

Source: MetroEngines Research; compiled by the HotelBank Editorial Team

This shape shows that demand in day-trip-heavy prefectures has settled into a “one night on Saturday, home on Sunday” pattern. Read the other way, Friday nights, Sunday nights and weekday nights leave clearly identifiable rooms to fill and the capacity to receive them. Looking only at annual average occupancy does not reveal where the conversion opportunity sits.

Evening Value 1: Dinner Adds ¥6,585 to ¥13,770 per Room to the Rate

In building a reason to stay, dinner is what feeds most directly into the rate. Taking only cases where the same property sells room-only, breakfast-included and two-meal plans simultaneously, a paired comparison (June-August 2026, double occupancy, all segments) gives the following per-room price differences from moving a breakfast-included plan to a two-meal plan.

Source: MetroEngines Research; compiled by the HotelBank Editorial Team

Table 5 | Dining scores are from a 13-category aggregation of guest reviews (most recent 24 months).

Source: MetroEngines Research / guest-review aggregation by the HotelBank Editorial Team

The largest is Hyogo at +¥13,770 (72 properties with a valid pair), followed by Nara at +¥10,632 (18 properties), Shiga at +¥9,790 (37) and Yamanashi at +¥9,325 (44). Where the breakfast premium (room-only to breakfast-included) sits in a +¥513 to +¥3,004 range, dinner is an order of magnitude larger. A single-prefecture deep dive using the same same-property pairing method is available in Ibaraki Breakfast Premium +5.8% vs Two-Meal +52.3% (July 2026). How the evening is designed feeds straight through to the per-room rate.

And the groundwork is there. Looking at within-prefecture medians of the “dining” category score in guest reviews (most recent 24 months), Saga at 4.51 (92 properties, 9,907 reviews), Nara at 4.50 (77 properties, 5,820 reviews) and Hyogo at 4.49 (328 properties, 40,696 reviews) all exceed the 47-prefecture median of 4.39. Mie at 4.45 (234 properties, 30,626 reviews), Gifu at 4.45 (221, 22,672) and Shimane at 4.44 (114, 14,101) do the same. Guests are already praising the evening meal, giving many of these prefectures a foundation to build on.

Evening Value 2: Put Plan Design on Friday, Sunday and Weekday Nights

As Table 4 showed, demand in the 20 prefectures concentrates on a single Saturday night. Saturday already runs at 90.3% estimated occupancy (average of actual dates in June and July 2026, as below), so adding a new plan there tends to displace existing demand rather than add to it. The targets for design are Sunday at 79.5%, Monday at 80.0% and Friday at 84.4%.

Three specific directions suggest themselves. First, designing for a Sunday night. Most day-trippers make their trip on either Saturday or Sunday. Staying Sunday and leaving Monday morning has a limited audience given that the next day is a working day, but it works for older and retired travellers and for remote workers. Given that Sunday’s listed price is 74% of Saturday’s, rather than discounting to drive demand it protects rate better to build in experiences that are hard to offer on a Saturday — private use, small-group dining, morning activities — and list them at a comparable price.

Second, pulling the stay forward to Friday night. Friday’s estimated occupancy of 84.4% (average of actual dates in June and July 2026) is 5.9pt below Saturday, and its listed price is 76% of Saturday’s. Extending a single Saturday night into two nights from Friday is not incremental acquisition but a longer stay from an existing guest, which makes it the lowest-cost option. Changing the dinner menu on the second night is one design that works here.

Third, making weekdays a destination. Day-trip-heavy prefectures are often around two hours from a major metropolitan area. If a weekday evening can offer something that can only be done in that place, it reaches the segment sitting between business and leisure demand. Aggregating published dinner-related plan names in Nara, two-meal plans account for 46%, kaiseki courses 24% and open-air hot springs 19%, making food and bathing the central appeal axes (N=472 plan names). There is still room to add elements that make the evening hours explicit.

Evening Value 3: Price on a Rate That Includes the Evening, Not a Discount to Saturday

What to avoid in pricing is filling quiet days with a discount against Saturday. Across the 20 prefectures, Sunday’s listed price is 74% of Saturday’s and Monday’s 69%, but in most cases that difference already reflects the demand gap, and cutting further is unlikely to drive conversion. Day-trippers stay away less because of price than because there is no reason to stay.

As Table 5 shows, a two-meal plan carries a real price difference of +¥6,585 to +¥13,770 per room. If a plan with dinner built in were designed at around the room-only Sunday level (¥18,962 across the 20 prefectures), it could be listed at a price band close to Saturday’s listed level (¥25,705). The idea is to change the product composition by day of week rather than the price.

Estimated settled ADR levels are also instructive. Among the 20 prefectures in Table 2, Hyogo (¥13,255), Nara (¥13,178), Yamanashi (¥11,749) and Gunma (¥12,001) sit in the ¥10,000s, while Ibaraki (¥6,951), Ehime (¥6,931) and Iwate (¥7,192) sit around ¥7,000. The latter group has a higher share of limited-service accommodation, with many properties that do not serve dinner. In those prefectures, building the evening experience together with nearby restaurants is likely more realistic than each property preparing dinner alone. This is where a DMO can play a role.

Note that ADR in this article is an estimate; cross-checked against property-level actuals disclosed by listed REITs, the median error is roughly 7%. Individual properties should base pricing decisions on their own actuals.

Saitama, Aichi and Chiba Need a Different Prescription — Heavy Day Trips, but Rooms Already Full

Three prefectures sit below the median conversion ratio while exceeding the 61.6% national occupancy average: Saitama (0.328, 66.7%), Aichi (0.593, 69.2%) and Chiba (0.696, 64.0%).

Saitama has the lowest conversion ratio in the country, but converting to the median would imply 4.57 million additional room-nights a year against 2.06 million vacant room-nights — 222%. The theoretical converted demand far exceeds existing vacancy, and the implied occupancy uplift of +73.9pt computes to over 100%. That points less to large room for demand conversion than to room stock that is simply thin relative to visit volume. Aichi, at 35% absorption and +10.8pt, is also more constrained than the 20 prefectures.

What these three share is being inside or immediately adjacent to a major metropolitan area, where the day-trip catchment structure itself is strong. The question here is less about designing evening content than about the volume and location of room supply. Even on the same “heavy day-trip” number, the course of action only separates once spare supply capacity is laid over it.

Properties Already Scoring Highly on the Evening Meal — Examples from the Conversion-Opportunity Prefectures

Finally, here are the properties within the 20 conversion-opportunity prefectures whose “dining” category score in guest reviews is particularly high, limited to properties with at least 30 reviews aggregated over the most recent 24 months. These are cases where the value of the evening stay actually shows up in the scores.

Source: Guest-review aggregation; HotelBank Editorial Team research

What lines up is mainly small ryokan, auberges and pensions with around 10 rooms. As Table 2 showed, room stock in the conversion-opportunity prefectures is by no means large, but properties already earning high scores around the evening meal exist within it. Whether a prefecture converts as a whole depends on whether this model can be spread across the area.

Summary

On the 2025 final figures, 45.8% of Japanese domestic trips are day trips, and whether a visitor stays or goes home makes a difference of ¥52,412 per person per trip. The thickness of this demand that goes home without staying varies 6.9-fold between prefectures, and the 20 prefectures meeting both conditions — conversion ratio below the median and occupancy below the national average — hold 7,114 properties and 317,787 rooms to receive it.

If those prefectures reached the median conversion ratio, the additional demand of 13.85 million room-nights a year would be just 19% of their existing 72.48 million vacant room-nights. It is a volume that can be absorbed without adding supply, which moves the question to demand-side design. By day of week, the night people stay is concentrated on Saturday alone, leaving around 10pt of occupancy headroom on Friday, Sunday and weekday nights. And dinner lands directly on rate, at +¥6,585 to +¥13,770 per room.

That said, in prefectures such as Nara and Saga the theoretical converted demand reaches 60-90% of vacancy, and room stock itself becomes the constraint. Even on the same “heavy day-trip” reading, the course of action separates only once spare supply capacity is laid over it. All figures in this article are desk estimates of an upper bound, not forecasts.

Related Reading

• Gifu 8.3% vs Wakayama 29.7%: Overnight Stay Rates in 20 Prefectures (Analysis on the common-standard visitor-arrival statistics, 2022 annual figures. The source, year and prefectures covered differ from this article.)
• Japan FY2025 Travel Spend ¥8.39tn: Growth Went Outside the Room Bill
• Domestic Demand H1 2026: Tottori & Ibaraki +27.7%, Kochi OCC +7.2pt
• Japan’s ¥2.83tn Hotel Room Revenue Pool: Top 5 Prefectures 48.0%
• In-Room Dining Ryokan TOP 30: 58.3% Still Offer It, ADR +91%
• June 2026 Occupancy 56.2%: Guest-Nights per Room in 47 Prefectures
• Ibaraki Breakfast Premium +5.8% vs Two-Meal +52.3% (July 2026)
• Inbound Share vs Weekday Gap: −0.13 Across Japan’s 47 Prefectures
• Inbound Stay Length Varies 2.0x: Tokyo 2.32 vs Yamanashi 1.16 Nights